Keynote Address at the e-Forum: MFPC Professionalism & Ethics Forum 2023

Speaker: Dr. Wong Huei Ching
Location: Virtual
Delivered: 8 November 2023

Full Speech:
Mr Andy Ng, President of the Malaysian Financial Planning Council
Mr Alvin Tan, President of Financial Planning Association of Malaysia
Mr John Chan, President of the Association of Financial Advisers
Mr Sean Lee, President of the Malaysian Financial Planner & Advisers Association
Mr. S. K. Samy, Chairman of the Ethics and Compliance Board of Malaysian Financial Planning Council

Ladies and gentlemen, good morning.

I am delighted to join you at the Professional & Ethics Forum 2023. Thank you, Malaysian Financial Planning Council (MFPC) for inviting me to deliver the keynote speech for this year’s virtual industry forum.
The theme “Ethical Considerations in Artificial Intelligence and Machine Learning” is especially relevant as technology rapidly evolves. This underscores MFPC’s commitment to conscientious and progressive stewardship in raising the bar within the financial planning profession.

The promise of Artificial Intelligence (AI) and Machine Learning (ML)

Ladies and gentlemen,
Artificial Intelligence (AI) and its sub-branch of Machine Learning (ML) is heralded as the next digital frontier given their transformative potential. They are able to automate tasks intelligently, extract insights or patterns from extensive amount of data and personalise outcomes.
Various publications have estimated that productivity gains and cost reductions arising from the adoption of AI can contribute up to USD 15 trillion to the global economy by 2030 to 2035.
In fact, we are already seeing the use of AI everyday. This ranges from Spotify’s personalised song playlist, to Netflix’s recommended watchlist to Google Map’s optimised driving routes from our home to the office.
The demand for personalisation is also revolutionising investments and advisory. AI holds the promise of powering sophisticated and risk-optimised investment strategies, as well as synthesising multi-dimensional information for personalised investment advise.
In the context of the Malaysian capital market, the SC’s 2021 survey revealed that 27% of 97 respondents were already adopting AI and ML, and a further 9% were in development stages. Of these respondents, fund management firms take the lead. Most are adopting AI and ML to automate KYC as well as customer risk monitoring. We also find that many are adopting AL and ML in investment analysis and portfolio
The adoption amongst small capital market players are also very encouraging. Early this year, the SC launched the Digital Innovation Fund (DIGID) to seed projects that will push the boundaries of innovation with technology. I am pleased to say that most of the 8 DIGID Award recipients todate are embarking on AI-powered innovations or paving the way towards AI adoption.
Generative AI (GenAI), made popular by ChatGPT, refers to AI that is capable of generating new content. While traditional AI is good at finding patterns in data, GenAI is good at using those patterns to create new and personalised content. For example, GenAI has the potential to generate customised investment recommendations, by synthesizing various client data and market data.

Responsible and ethical use of AI and ML

Ladies and gentlemen,
The effectiveness and reliability of any technological tools, from simple rule-based algorithms to sophisticated AI and ML, fundamentally hinges on two pillars: the quality of data and the appropriateness of the models used. If either is lacking, the technology's reliability is compromised.
For rule-based algorithms, the principle 'garbage in, garbage out' holds true; subpar data leads to unreliable outcomes, and models that do not fit their purpose will fail to perform as intended.
When it comes to AI and ML, the stakes are higher, given the scale of its ability. These systems use large volume of data sets to learn and make decisions. If the data is of poor quality or the models are ill-suited, the errors can echo and grow within these systems.
In the case of GenAI, the model can perpetuate biased or incorrect content based on flawed or incorrect inputs. Imagine the market impact when GenAI perpetuates incorrect investment recommendations at scale based on misinformation.
This scenario highlights the role of human touch in the era of AI. While AI offers deep insights, the oversight of ethical advisors remain pertinent, to ensure that decisions are made with the clients' best interests at heart, always safeguarding their financial goals.
In short, the responsible use AI and ML entails managing the inherent risks. These are risks related to the model, such as risk of inaccuracy, biasness, ethical concerns and lack of explainability. There are also risk related to data integrity, such as poor data quality, data manipulation, labelling errors and others that impacts the inputs to AI models.
The SC’s Technology Risk Management Guidelines, issued in August 2023, includes guiding principles on the adoption of AL and ML by capital market entities to manage these inherent risks. These principles align closely with the primary themes underpinning the ethical use of AI and ML techniques identified by IOSCO Fintech Network .
Broadly, the following four (4) guiding principles should be implemented in proportionality when considering the ethical design, development, and deployment of AI and ML in the financial planning industry.

Accountability
Firstly, the need for accountability over the development and use of AI.
A governance framework with clear AI objectives, well defined roles, responsibilities and lines of authority can ensure that the development of AI is fit for purpose.
Company leaders or senior management play a key role to guide the responsible use of AI. This entails enabling innovation with risk management measures in place.
Clear communication about how and why AI and ML are used can also build trust with the clients.

Transparency and Explainability
Secondly, transparency and explainability of the AI model, is pertinent to reduce risk of opaque AI models.
This means documenting decisions by the AI with clear explanations on both its method or process and outcome.
Such clarity is key to building trust and ensuring that all actions taken by AI can be communicated in simple and accessible terms.

Fairness and Non-Discrimination
Thirdly, the need for AI models that operate without prejudice, and offer consistent and unbiased outcomes to all individuals, regardless of background.
To do so, both data and AI model will need to be reviewed regularly to minimise unintentional bias, and to ensure that the use of AI is in alignment with the firm’s ethical values.
In other words, AI-driven decisions should be held to the same ethical standards as human-driven decisions.

Practical Accuracy and Reliability
Lastly, investors will expect your FPs to work with reliable AI-driven advice. This means having reliable AI models and accurate data.
To achieve this, FP firms will need to have in place appropriate controls for rigorous testing and data quality, as well as have necessary security measures for data protection.

Conclusion

Ladies and gentlemen,
As FPs and advisors integrate the use of AI into the profession, it is a collective responsibility to ensure that every decision made with AI and ML aligns with your clients' best interests – be it in wealth preservation, protection, accumulation, or distribution.
As we navigate the AL and ML frontier, there is much invaluable experiences and insights to be shared and gained from each other. With the line-up of speakers today, I am confident that everyone will be in for a wealth of knowledge and enlightening discussions.
I wish you all a fruitful and engaging forum ahead. Thank you.